An Improved DFC-Based Cooperative Spectrum Sensing for Cognitive Radio in the Presence of Malicious Users

Abstract

In cognitive radio networks (CRNs), cooperative spectrum sensing (CSS) can effectively improve sensing performance by taking advantage of spatial diversity of participating second users (SUs). However, the spectrum sensing data falsifying attack (SSDFA) caused by malicious second users (MSUs) severely degrades the performance of CSS networks. In this paper, to defend against the SSDFA of MSUs, we propose a new CSS mechanism based on distributed fusion centers and overlapping clusters (DFCs-OCs). With this new architecture, our proposed linear combining CSS algorithm improves the CSS performance by mitigating the effects of MSUs on decision making through updating the reputations of SUs and OCs. Extensive simulation results prove that the proposed CSS scheme performs far better than related recent studies under both high and low signal-to-noise ratio conditions in CRNs, where MSUs occupy nearly half of all SUs.